Smart Railway TFDS Solution
Precisely identify faults and streamline operations with the
Huawei Pangu Railway Model.
Productos, soluciones and servicios para los negocios
Precisely identify faults and streamline operations with the
Huawei Pangu Railway Model.
A conventional trouble of moving freight car detection sytem(TFDS) uses trackside high-speed cameras to capture images of freight car components. Images are then analyzed by TFDS inspectors, one by one, to identify potential train faults and generate alarms accordingly.
But, with hundreds of millions of images generated every day, enormous pressure is placed on inspectors, resulting in all too frequently missed faults and alarms that are never raised. That's a big problem. So Artificial Intelligence (AI) and similarly advanced technologies are urgently needed to improve the intelligence of TFDSs, accelerate fault analysis, and relieve the pressure placed on inspectors.
Huawei's Smart Railway TFDS Solution deploys the Huawei Pangu Railway Model as the foundation for pre-trained models, integrating AI into the entire TFDS identification process. This effectively filters out over 95% of fault-free images, to precisely identify over 430 common faults on 67 types of freight cars,the comprehensive fault identification rate is greater than 99.3%. Greatly improve operation efficiency.
Gain the Edge
Precise Identification
Covering both human-machine collaboration and through-type operation scenarios, the solution identifies over 430 common faults with a comprehensive fault identification rate over 99.3%, ensuring near-zero missed critical faults.
Increased Operational Efficiency
The solution intelligently filters out over 95% of fault-free images, significantly reducing manual review workloads and increasing operational efficiency by 200%.
Uninterrupted Teaching
• Full coverage of high-bandwidth wireless networks for teaching and learning anytime, anywhere
• Multi-service access without frame freezing or disconnection, ensuring teaching continuity
Products
Atlas 800 Inference Server (Model: 3000)
Powered by the Ascend processor, the Atlas 800 inference server (model: 3000) supports up to 8 Atlas 300I inference cards to provide powerful real-time inference.
Más información
CloudEngine S5731-S Series Switches
• GE electrical/GE optical /2.5GE optical downlink, 4*10GE SFP+ uplink ports
• Switching capacity of 672 Gbit/s
• Super Virtual Fabric for simplified network O&M
• One network for multiple purposes through VXLAN-based network virtualization
Más información